In patients with sinoatrial disease, unexpected atrial flutter (Af) or fibrillation (AF) is a common problem during implantation of atrial-based pacing systems. As an alternative approach to blind atrial lead placement, lead positioning could be optimized by atrial electrogram mapping. It was the object of this study to evaluate if atrial lead implantation according to this approach and during ongoing arrhythmia is reasonable or if it should be postponed until restoration of sinus rhythm (SR). Twenty-nine consecutive patients (group I) with sick sinus syndrome received a dual-chamber pacemaker during an episode of Af (n = 11) or AF (n = 18). All but two atrial leads were of the screw-in type and had bipolar sensing. Atrial lead position was optimized by mapping the electrogram under fluoroscopy to find locations with high potential amplitudes. The patients were followed for 15.1 +/- 9.8 months, and atrial sensing threshold (AST), atrial pulse width threshold (PWT) at 2.0 V, the pacing mode programmed, and the clinical outcome (OUT) were recorded. The control group consisted of 30 patients (group II) who equally had a history of AF or Af, but were in SR during implantation. The atrial peak-to-peak potential (APEAK) after final lead placement was lower for AF (median value 2.5 mV, lower-upper quartile: 1.7-3.1 mV) as compared to Af (3.8 mV, 2.7-4.9 mV, P < 0.05) and SR (4.1 mV, 3.3-6.2 mV, P < 0.001). There was a correlation (P < 0.01) between APEAK during Af/AF and the postoperative AST immediately after restoration of SR. No lead in any group had to be corrected due to improper sensing in the postoperative course. Median chronic AST was 2.8 mV (2.0-4.0 mV) in group I and 4.0 mV (2.8-4.0 mV) in group II. Median chronic PWT at 2.0 V was 0.15 ms (0.12-0.26 ms) in group I and 0.15 ms (0.09-0.20 ms) in group II. There was no significant difference in chronic AST and PWT between both groups. All but two patients in group I preserved SR as the basic rhythm. A stable SR was observed in 10 of 29 patients, intermittent Af/AF was documented in 17 of 29 patients, seven of whom were asymptomatic. There was no significant difference in OUT between group I and II. Hence, sinus rhythm is not a prerequisite of atrial lead implantation. Mapping the Af or AF waves appears to be useful to guide lead placement and to achieve sufficient sensing and pacing conditions after conversion to sinus rhythm.